RE: [cp-patches] RFC: PR Classpath/28664 - GNU MP based BigIntegertake2 (compressed)

2006-10-08 Thread Jeroen Frijters
Hi Raif, The synchronization is still not completely right. If a method operates on two different BigIntegers, they both need to be protected from asynchronous finalization. Note also that when you synchronize on two different objects, you need to protect against deadlock. Also, could you please

Re: [cp-patches] RFC: PR Classpath/28664 - GNU MP based BigIntegertake2 (compressed)

2006-10-08 Thread Raif S. Naffah
hello Jeroen, On Sunday 08 October 2006 22:44, Jeroen Frijters wrote: Hi Raif, The synchronization is still not completely right. If a method operates on two different BigIntegers, they both need to be protected from asynchronous finalization. Note also that when you synchronize on two

[cp-patches] Re: preparation of VMPlainDatagramSocketImpl removal

2006-10-08 Thread Casey Marshall
Robert Schuster wrote: The other thing that is used from VMPlainDatagramSocketImpl is the connect() method. Is it possible for DGRAM sockets to just use VMChannel.connect instead? Yes. The original just called the common method _javanet_connect, anyway. Great. However there is just one

[cp-patches] RFC: VMPlainDatagramImpl removal

2006-10-08 Thread Robert Schuster
Hi, this patch removes the above mentioned class and implements their functions elsewhere. DatagramImpl.connect() has a different exception clause than SocketImpl.connect() and fixing that required some changes in method signatures and the native implementation of VMChannel. These changes are